Canon EOS 50D Sharpness. Am I paranoid/Opinions please!

Chief

Hey folks, I now know this topic has been bandied about on forums all over the net but to be honest I never knew there was so much controversy on the net about this issue. Personally I always thought my images with my 24-105mm lens were a bit blurry but I always put it down to the fact that my lens had a wee dent in it. I finally got the chance to sit down with a photo journalist m8 of mine and do a couple of tests today. He has a 1D - MK III and we decided to try my lens on his and mine consecutively. I was shocked and surprised to be honest with the results.

1D - MK III cropped 100%. 85mm f4 iso400

pretty damn sharp...

50D cropped 100%. 47mm iso100

blur city

! mine super blurry... his sharp as hell with my lens... pores.. hair.. lips.. sharp.
Now to be honest I thought with the digic 4 and higher mega pixel mine would be sharper. Maybe the wrong assumption?

So i read a bit on dpreview forums etc and have seen some SUPER sharp shots and I've read how the camera can be adjusted for sharpness etc with each individual lens. So I'm gonna try that tommorow when I get the chance.

I also conducted one more personal test using Aperture Priority and f4 then f8. I would just like to know do my fellow Canon users think I'm paranoid, too fussy or am I barking up the right tree? Do I need a wee service by Canon? Well out of warranty now unfortunately. Here are my test images

50D f4 105mm iso1600 100% Crop


and...

50D f8 105mm iso1600 100% Crop


Thanks for your time and advice folks...

Borderfox

post some up with a link to the full size shots

charybdis

Your tests are nowhere near scientific enough. If you really care, go buy a resolution chart and rigorously measure how many lp/mm you get at different focal lengths, aperture values, and focus distances then compare it to a similarly rigorously generated data set for another 50D and see if there's actually a slight problem with your sensor.

It's also fairly pointless to compare a 100% crop of an image generated by your "m8"'s 8 megapixel APS-H sensor to a 100% crop of an image generated by your 15 megapixel APS-C sensor, the dot pitch of your camera is far higher than his and as such is far harder for any lens to out-resolve.

Anyway, I doubt it's the sensor. If anything I'd say the slight blurring and chromatic aberration suggest imperfect focus. Again, it's hard to tell with these uncontrolled tests.

Just go and use your camera and don't worry about how the images it produces look at 100%.

Chief

Thanks for the input charybdis. I know the tests are very unscientific but when you see a bunch of shots from another camera using your lens which are sharp on his camera lcd and the same photographer takes almost identical shots with your lens and your camera and they are blurry on the lcd, you can't help but worry.

Would you not worry if your images were always blurry at 100%?

I'm gonna fork our for a resolution test card and check with other 50D users about their test data. I'll try the fine adjustments on the camera itself today also. Thanks again for the advice charybdis

.Longshanks.

I thought the 50D has a micro focus adjustment feature specifically for this kinda thing....?

Chief

Longshanks100 wrote: »

I thought the 50D has a micro focus adjustment feature specifically for this kinda thing....?

.

Just discovered all about it yesterday Longshanks, gonna give it a go today
When I got the camera from pixmania it came with every manual in every language except English so I kind of just moved on from using my Canon D60 to the 50D in a similar way not knowing much of the new features that I've discovered this last while.

ttcomet

You can get a free test chart here. Just download it and print it out then stick it on to some stiff cardboard.

Chief

thanks ttcomet!

Bosshogg

Do you know that these cameras come with a ton of settings for how the exposure is processed? You can adjust the sharpness too. Also in post processing with Digital Photo Professional they get another going over. It's worth having a dickyabout with all that.

Heebie

Lenses have something called "resolving power".. it's really a measure of how sharply they are able to focus light over.. (this focus not being the same thing as focusing you're used to.)

If your image sensor's resolution exceeds the resolving power of the lens.. you'll see the types of things you're seeing.

On your friend's camera, the resolving power was high enough that individual pixels in the sensor were seeing sharp lines.. whereas on your sensor which has 1.5x the sensor-density.. the same part of the image focused by the lends is spread across multiple pixels, resulting in something very similar to looking at a 320x200 pixel image on a modern 1920x1280 (HD) display.. it'll look out of focus and a bit blotchy. In an auto-focus system that relies on the image sensor, this could also cause actual focusing problems. (You could try re-running this test using manual focus & making sure that you're REALLY happy with the sharpness of what you see through the viewfinder before you shoot.)

The exposure is also very different between the two photos, and the angle of the person's face in the photo is different. You should try a more controlled test with a subject that doesn't move, on the same tripod, and under the same lighting.

You could also have been using different auto-focus modes on the two cameras. (again, using manual focus would fix this.) The first photo looks like it was probably shot using some focus method that was centered on the person's face. (not knowing the MK III, I have no idea how many focus-points the AF system can choose.) The second looks like it might have been looking for an overall average focus for the photo, rather than "looking" for a person's face, or being focused on a specific point.
If you're going to use auto-focus, you'll want to use whatever controls you have to make sure that both cameras are concentrating on the same position in the photo for their focus, and are using the same modes.

The exposures being different could be the result of different exposure modes, and/or different aperture settings. Even using the most similar modes on two different cameras could cause very different results in what the automatic systems "choose" to use.

Maybe re-try this with a good tripod, a stationary subject (perhaps the really excellent-looking focus chart that someone else refers to in a post here as ONE of the subjects.) , using fully manual exposure and focus on both cameras.. and shoot in RAW mode so you're seeing exactly what's hitting the image sensors, and not looking at the post-processed JPG version of the picture that the internal processing chip came up with. That should level the playing field between the two cameras as much as humanly possible so you can really be looking at ONLY the differences between the lens' perceived quality of focus (or lack thereof as the case may be.)

I would also try this same experiment using a different lens as a control.

If you find that both lenses are doing the same thing..then there's something wrong with your camera. (probably a physical flaw of some kind.) If it's just that one lens, then it could be that that lens doesn't have very high resolving power, and that the sensor density of your camera is too high for that lens. [this is MUCH more likely in lenses that have plastic elements within them than with lenses that are all glass elements.. you should be able to find the construction of the particular lens online somewhere.)

I hope this was helpful and that I don't just sound like a prick!